On the Trilobite Genus Celmus Angelin, 1854

By Valdar Jaanusson

ABSTRACT.-The trilobite species Ceimus granUtatus ANGELIN, 1854 is redescribed. Following FR. ScHMIDT (1907) the genus Crotalurus VoLBORTH, 1858, is considered as a junior subjec tive synonym of Celmus ANGELIN, 1854· A new family Celmidae is erected with Celmus as the sole representative. The terms eu-ptychopariid and steno-ptychopariid type of ventraJ cephalic sutures are introduced to denote a broad and a narrow rostraJ shield, respectively, separated by sutures. A new subfamily Mesotaphrasinae is erected for certain genera with levisellid type of facial suture including Mesotaphrasis, Chomatopyge and, provisionally, Clelandia.

In 1854 ANGELIN (pp. 23-24, PI. XVII, fig. 8-8 a) erected a new trilobite genus and species narned Celmus granulatus derived from strata regianis C, Skarpåsen, Östergötland. He illustrated and briefly described the cephalon and thorax whereas the pygidium was apparently unknown to him. The genus was left unplaced in family. In 1858 VoLBORTH (pp. 126-133, PI. XII, fig. 1-5) described a trilobite from the Lower Ordavieian limestones at Pavlovsk, Ingermanland, believing it to represent a new genus and species which he narned Crotalurus bar randei. His thorough description comprised the whole carapace with the exception of the hypostoma. The genus was left unplaced in family. As far as can be judged from the references in his paper the monograph of Angelin (1854) had not been available to him. The genus Crotalurus was included in the family no. XV (Cheiruridae) by BARRANDE (1872, p. XXX). FR. ScHMIDT (1881, pp. I 19-120) excluded it, however, from the Cheiruridae on account of its opisthoparian facial suture hut left open the taxonomical position of this genus. Subsequently FR. ScHMIDT (1907, pp. IS-I6, PI. I, figs. 9, 10, u ) published a redescrip tion of VoLBORTH's material of Crotalurus barrandei VoLB. From a study of photographs of the type specimens of Celmus granulatus he conducled that Celmus and Crotalurus are subjective synonyms, and that also the type species of these genera appear to be synonymous. Nevertheless, he continued to use VoLBORTH's name for the genus and species painting out that VoL BORTH's description was more complete, and that the generic name Crotalurus had been more widely used in the literature than Celmus. In this paper

3-553271 Bull. of Geol. Vol. XXXVI VALDAR JAANUSSON

FR. ScHMIDT again included Crotalurus tentatively in Cheiruridae chiefly because no better place for this genus could be found ( cf. also STUBBLEFIELD I936, p. 4I6). WEBER (I932, pp. I22-I23) included Crotalurus in Proetidae on account of the general resemblance of the glabella in Crotalurus and Cyphaspis. He also described the new species Crotalurus? strigatus from Tamdy River, Karatau Range, bu t pointed out in a footnote (p. I22, foot note I) that this species no doubt belongs to Glaphurina erected by ULRICH (I930) in a paper which became accessible to WEBER first after his manu script had already been submitted for publication. ÖPIK (I937, p. 92) sug gested that Celmus (incl. Crotalurus) should probably be placed into a new opisthoparian family. Recently TJERNVIK (I952, pp. 6I, 64) recorded Crotalurus n. sp. from the Lower Planilimbata limestone of Närke, and BOHLIN (I955, p. II5) Celmus granulatus from the Raniceps limestone of Öland, and (I955, pp. I27-I28) Celmus aff. granulatus from the Gigas limestone of Öland. In connection with the preparation of the trilobite part of the Treatise of Invertebrate Paleontology it appeared necessary to settie the taxonornie position of the genus Celmus. For this purpose the present writer undertook a study of all available material of Celmus from Sweden, including the cotypes of ANGELIN. In the present paper a rede8cription of the type species of Celmus is given tagether with a discussion on the taxonornie position of the genus. The pygidium, previously unknown from Sweden, and the hypos toma, hitherto not known in this genus, are also included in the description. The material studied belongs to the following museums: State Museum of the Natural History (Naturhistoriska Riksmuseet) in Stockholm (RM), Museum of the GeologicaJ Survey of Sweden (SGU), and Museum of the Palaeontologieal Institution of the Univ. of Uppsala (UM).

Remarks on the Terminology In the literature on trilobites written in English language the middle region of thorax and pygidium is frequently termed "axis", and the furrows which limit this region laterally-the "axial furrows". These terms have been u sed by SALTER (I 864, etc.) who was followed in this respect by WARBCRG (I925) in her critical revision of the terminology of the trilobite carapaee. The fact that WARBURG herself in her later papers ( e.g. I939, p. 6) found these terms to be inadequate and replaced them by the old terms "raehis" and "dorsal furrow" is, however, less known. The word "axis" means a line and its use for an area of a body is not quite happy. The term "axial lobe" u sed by WALCOTT (I 908 and later p apers) is a more suitable term if the use of the word "axis" for the middle region of the trilobite Clrapaee is preferred. For this region DALMAN (I 827) had introdueed the term "rhachis" (in English usually "raehis") and as this paper of DALMAN may be regarded as the foundation of the scientific terminology of the ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854 37 trilobite carapace, his terms may reasonably be retained. The term "axial furrow" in the sence of "sulcus dorsalis" of DALMAN (1827) is, however, from the point of view of zoological terminology, entirely misleading and ought to be dropped. DALMAN's term "dorsal furrow" has been widely accepted (e.g. BARRANDE 1852 ["sillon dorsaux"], RrcHTER 1912 ["Riicken furche"], WARBURG 1939, Howeli et al. 1947, to mention only a few out standing discussions on the terminology of the trilobite carapace), and is used also in the present paper. For brevity as well as darity symbols have been used in the present paper to denote the lateral glabellar lobes and furrows. The present writer completely agrees with WARBUR G (1939, p. 5) in considering unnatmal the numbering of glabellar furrows and lobes from behind forwards. On the other hand, as pointed out by several writers (most recently by WHITTING TON & EviTT 1954, p. 13), this method of numbering provides at present the only possibility assigning the same figures to homologous furrows and lobes. For purely practical reasons, therefore, the furrows and lobes are, following RICHTER & RICHTER (1940, p. 17, fig. 3 A), numbered from back to front. The lobes are lettered "L" (after lobus) and the furrows "S" (after sulcus). In the terminology used previously by BARRANDE (1852, Pl. 1, fig. r), RICHTER & RICHTER (1926, p. 126, fig. 14), and the present writer (1954, p. 55r, fig. 5) the basal (posterior) lobe (furrow) earresponds to Lr (Sr), the middle lobe (furrow) to L2 (S2), and the anterior lobe (furrow) to L3 (S3). These figures in combination with letters L and S ought to be used only if the method of numbering is that applied in the present paper, i. e. from back to front. In this case the letters "p" ( = posterior) or "a" (=an terior) after the figures giving the direction of numbering, as suggested by WARBURG (1939, p. s), are superfluous. The terminology of the different types of ventral cephalic sutures used in the present paper earresponds to that of RASETTI (1952). The type with a rostral shield separated by sutures is accordingly called the ptychopariid type. Within this type it is, however, advantageous to distinguish two sub types. In the present paper the genera with a broad rostral shield are spoken of as the eu-ptychopariid type (as in Ptychopariidae, Illaenidae, Scutellidae, Cheiruridae etc.). To the genera with a narrow rostral plate (as in Encri nuridae, Celmus, Catillicephala, Hystricuridae, Dimeropyge, Dimeropygiella, certain Bathyuridae etc.) the term steno-ptychopariid is applied. The typ e of ventral cephalic sutures where the anterior branches of the facial suture are fused, and no median suture is developed (as in Phacopidae, Nileidae, Leviselfa, Glaphurus, Mesotaphrasis etc.), is called, following RASETTI, the levisellid type, and, when a median suture is developed, the asaphid type. The different types of ventral cephalic sutures referred to in this paper are illustrated in fig. 1. VALDAR JAANUSSON

A c

Fig. r. Schematic illustrations of the different types of ventraJ cephalic sutures referred to in the present paper. A-eu-ptychopariid type; B-steno-ptychopariid type; C-asaphid type; D-levisellid type. r-rostral shield; m-median suture.

Fam. Celmidae n. fam.

DIAGNOSIS.-Small, apparently opisthoparian trilobites. Cephalon compara tively large, surrounded by a distinct convex horder. Glabella slightly nar rowing anteriorly, with two to three pairs of lateral glabellar furrows. L1 comparatively long and well-defined. Occipital furrow present. Rostral shield short (sag.) and narrow (tr.), trapezoidal. Hypostoma with rounded posterior margin, with small pointed posterior wings, and apparently narrow (tr.) and triangular anterior wings. Posterior lobe of the hypostoma well-defined, fair! y convex. 12 thoracic segments in the typ e genus; the width ( tr.) of the segments decreases considerably in posterior direction. Thoracic pleurae of the ridged type, without pleural furrows and facets. Pygidium very small, trapezoidal, apparently consisting of only one tergite; i t has no rachis hut two crescent-shaped elevations near the anterior margin. The type genus has a tuberculate ornamentation. DrscusSION.-The type genus being so far the only genus referable to this family the above diagnosis largely coincides with that of the genus Celmus. As pointed out by FR. ScHMIDT (1907) the genus Crotalurus VoL BORTH, 1858 is a junior subjective synonym of Celmus ANGELIN, 1854. In order to solve the question about the taxonomical position of the genus Celmus a large number of genera in some way or other similar to it has to be considered. Of some of these gen�ra available material has also been studied. As a result of the comparative examination of the similar genera some additional taxonomical problems arose in which the present writer is not quite in accord with current classification. ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854 39

In searching after the probable affinities of Celmus the following families must be considered on the first hand: Glaphuridae, Dimeropygidae, Hystri curidae, Catillicephalidae, and Cheiruridae. Several cranidia and librigenae of Glaphurus pustulatus (WALCOTT) were exaroined by the present writer (coli. H. B. WHITTINGTON 1950 and kindly presented by him to UM). This genus has a facial suture of the levisellid type (ULRICH 1930, p. 43, Pl. 8, fig. 15). The preglabellar field is Iong, the posterior lateral glabellar lobe bicomposite (L1 + Lz). The number of thoracic segments amounts to 10; distinct pleural furrows are present. Pygidium small with a Iong, broad, and strongly convex rachis composed of a small number of segments. Of Glaphurina (cf. ULRICH 1930) only cranidia are known so far. Also in this genus the posterior glabellar lobe is evidently bicomposite (Lr + Lz), hut there is no preglabellar field, the anterior border of the cephalon being in direct contact with the glabella. Both Glaphurus and Glaphurina were originally included in Telephidae (ULRICH 1930) hut later separated by HuPE (1953) as a special family Glaphuridae. Having exaroined material of both Glaphurus and Tel?phus the present writer agrees with HuPE (1953) in considering the former as belonging to a separate family. In fact the few points of similarity between these genera seem to be only superficial, and there does not seem to be sufficient evidence for a relationship between Glaphurus and Telephidae + Komaspididae to justify even their inclusion into the same superfamily, as suggested by HUPE (1953). The eyes in Glaphurus are small, fixigenae and preglabellar field broad and of quite different type compared with those in telephids or komaspidids. As the development of these characters certainly is more primi tive in Glaphurus than in telephids and komaspiclids it is improbable that this Chazyan genus should have arisen from the telephid-komaspidid stock which is fully developed already in the Upper Cambrian. Its relationship lies more probably somewhere in the vicinity of catillicephalids and related groups. The reference of Glaphurina to Glaphuridae is somewhat uncertain at present and cannot be definitely ascertained until other parts of the carapace are known. The cranidium of Celmus exhibits a certain externa! resemblance to Gla phurina. In the later genus, however, the posterior lateral glabellar lobe is bicomposite (L1 + Lz) whereas in Celmus it is formed by Lr only. The specimens recorded by TJERNVIK (1952) as Crotalurus n. s p., and apparently also Crotalurus? strigatus WEBER have a bicomposite posterior lateral gla bellar lobe, and probably belong to Glaphuridae. Glaphurus possesses a levisellid type of facial suture, and differs widely from Celmus also in other characters. There does not seem, therefore, to exist evidence of any doser relationship between the latter genus and Glaphuridae. There exists a number of opisthoparian genera which possess a steno ptychopariid type of rostral shield resembling that of Celmus. The following 40 VALDAR JAANUSSON genera have to be considered: Dimeropyge (cf. WHITTINGTON & EvrTT 1954), Dimeropygiella ( cf. Ross 1953), Pseudohystricurus (cf. Ross 1953), Hystricurus (cf. HrNTZE 1953), and Catillicephala (cf. RASETTI 1952, 1954). Dimeropyge, Dimeropygiella, and Toernquistia, tagether with the new genera Mesotaphrasis and Chomatopyge were included by WHITTINGTON & EvrTT (1954) in the family Dimeropygidae HuP.E, 1953. Dimeropyge and Dimero pygiella have a rostral shield of the steno-ptychopariid type, and the existence of a similar rostral shield has been suggested also for Toernquistia (WHIT TINGTON & EviTT 1954, p. 35). Mesotaphrasis and Chomatopyge have a facial suture of levisellid type. The inclusion of the latter genera in a separate family was considered by WHITTINGTON & EviTT (1954, p. 35) but the similarities with Dimeropyge, especially regarding the pygidium, were found by them to outweigh the differences in the development of the facial suture. They pointed out, however, that the inclusion into the same family of genera with such different development of the facial suture is a samewhat unusual course. In the present writer's opinion the genera with a facial suture of levisellid type should be included at least in a separate subfamily. Certainly, several cases are known where the development of the foremost part of the facial suture varies even in very closely related groups. In Dikelocephalidae, for instance, forms occur both with a median s uture (asa phi d type) as well as without it (levisellid type) (ULRICH & RESSER 1930; RASETTI 1952), and this family seems thus to be rather unstable in this respect. In Dimeropygidae as defined by WHITTINGTON & EvrTT (1954), however, the change from e.g. Dimeropyge to Mesotaphrasis implies first the loss of the rostral shield, i.e. a probable development from the ptychopariid to the asaphid type of facial suture, and only subsequently the loss of the median suture. In the light of the evidence from other trilobites the present writer considers this diffe rence in the development too large to allow the inclusion of both different types of facial suture in one subfamily, at least as long as an intermediate type of facial suture is unknown. Moreover, there exist also other important differences between Dimeropyge and the genera with a facial suture of levisellid type included in Dimeropygidae. In Dimeropyge the thorax is without pleural furrows; the dorsal side of the thoracic pleura is strongly convex, and narrow anterior and posterior flanges are developed (cf. WHITTINGTON & EvrTT 1954, p. 49, Fig. 9; the same construction of the thoracic pleura has been observed by the present writer also on detached thoracic segments of Dimeropyge parvula [THORSLUND]). In Mesotaphrasis a distinct diagonal pleural furrow is present, the pleurae being rather flat and on the whole similar to those in proetids. In the present writer's opinion the differences between Dimeropyge and Mesotaphrasis are important enough to justify their inclusion into separate subfamilies. For this reason the new subfamily Mesotaphrasinag is proposed here for Mesotaphrasis and related genera ( cf. p. 47). ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854- fl

The resemblance between Toernquistia and certain hystricurids (Parahys tricurus, Pseudohystricurus, Hystricurus) was pointed out by Dr. Ross in a personal communication to WHITTINGTON & EviTT (1954, pp. 35-36). The suggestion was discussed by the latter hut the relationship of this genus to hystricurids found to be uncertain at that time. Since then our knowledge of these hystricurid genera has been widened by the p apers of R oss ( r 9 53) and HINTZE (I953). HINTZE (I953, p. I64, PI. VI) described and figured the rostral shield in Hystricurus, and Ross (I953, p. 64I, PI. 63, fig. I7) pointed to the possible existence of a similar rostral shield in Pseudohys tricurus. In both cases the rostral shield is of the steno-ptychopariid type. In Hystricurus the thoracic pleurae possess pleural furrows (Ross I95I, PI. I4, fig. 27; HINTZE I953, PI. VI, fig. I b, I c), hut the published figures give no quite clear picture of the exact construction of the pleura. The thorax and pygidium of Toernquistia (REED I904, PI. XII, figs. 6, 7) seem to be rather similar to those of Hystricurus, and differ in certain important charac ters from those of Dimeropyge. In the present writer's opinion the known features of Toernquistia in fact resemble more those of hystricurids than those of Dimeropyge. On the other hand, our present knowledge of both hystricurids and dimeropygids does not exclude the possibility that both groups of genera are very closely related. Celmus differs from the hystricurids and Dimeropyginae by the presence of a distinct long LI, different construction of the thoracic pleurae, and the rather different pygidium. The details of the shape of the rostral shield are also different hut are probably of minor importance. In the construction of the thoracic pleurae Dimeropyge is samewhat similar to Celmus, princi pally by the absence of pleural furrows. In the former genus, however, there is a weil developed triangular facet, the doublure is much narrower, and the flanges are only weakly developed. On account of these rather consider able differences Celmus cannot be included in either Hystricuridae or Dimero pygidae. In the development of the rostraJ shield the genus Catillicephala is very similar to Celmus, and also other features, inter alia the shape of LI of some Catillicephalidae as defined by RASETTI (I954), show some resemblance to Celmus. The construction of the thoracic pleura and the pygidium of the latter genus, as weil as some other features, are, however, too different to permit the inclusion of Celmus inta Catillicephalidae. Moreover, the latter family as defined by RASETTI (I954) may not be homogeneous, especially with regard to the fairly great variation in this family as to the development of the ventral cephalic sutures. FR. SCHMIDT ( I907) pointed out that the thoracic pleurae of Celmus have the same construction as in most cheirurids, and mainly on this account placed this genus in Cheiruridae. Cheirurids are, however, proparian, have a rostraJ shield of the eu-ptychopariid type, and a differently shaped py- 4�2��- --··· ------VALDAR JAANUSSON gidium. In this case the similarity in the construction of the thoracic pleura seems to be due to a convergent development. Thus by its taxonomically important characters the genus Celmus differs clearly from other similar groups. The shape of its rostral shield, the con struetian of the thoracic pleura, and the peculiar pygidium place it in a rather isolated position among the similar groups, and its relationship remains rather uncertain. For this reason the new family Celmidae is erected for the genus Celmus.

Gen. Celmus ANGELIN, 1854

Syn. Crotalurus VoLBORTH, I858

TYPE SPECIEs (by monotypy).-Ctlmus granulatus ANGELIN, I854· This being the only genus known so far of the family the diagnosis of the genus coincides largely with that of the family (cf. p. 38).

Celmus granulatus ANGELIN, 1854 PI. I, figs. r-7; text fig. 2 A-D r 854 Gelmus granulatus n. sp.-ANGELIN, p. 24, PI. XVII, figs. 8-8 a. cf. r8s8 Grotalurus barrandei n. sp.-VoLBORTH, pp. rz6-r33, PI. XII, figs. r-s. non I907 Grotalurus barrandei VoLBORTH-FR. ScHMIDT, pp. IS-I6, PI. I, figs. 9, IO, I I. non I955 Gelmus granulatus ANGELIN-BOHLIN, p. I 15 ( � Gelmus sp. indet.).

LECTOTYPE (designated herein).-A carapace without pygidium RM no. Ar. I7989, one of the cotypes of ANGELIN (I854), and used by him in the preparation of the reconstruction on PI. XVII, fig. 8. Figured in the present paper on PI. I, figs. I-3· TYPE LoCALITY.-Skarpåsen, Östergötland. TYPE STRATUM.-Lower Ordovician, probably Expansus limestone. DIAGNOSIS.-No proper diagnosis can be given as long as only the type species is sufficiently well known. MATERIAL.-Östergötland-2 enrolled carapaces without pygidium ( cotypes); Närke-I carapace, IO cranidia, IO librigenae, 4 hypostomata, 4 incomplete thoraces, 3 pygidia. DESCRIPTION.-Carapace broadest at cephalon and narrowest at pygidium. The width of the pygidium is slightly less than that of the frontal lobe of the glabella. Glabella more or less rectangular in outline, slightly narrowing forwards, greatest width at LI, the anterior margin of the frontal lo be being only slightly convex. In dorsal view the length of the glabella (excl. the occipital ring) about equal to its greatest width. In lateral view the foremost part of the frontal lobe is stecply inclined in relation to the hindmost part of the glabella and is concealed in strictly dorsal view (cf. PI. I, figs. I and 3). ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854 43

Sr rather strongly developed, decpest in its middle portion, becoming shallower in posterior as weil as antero-lateral direction. In posterior direction Sr becomes almost obsolete before reaching the occipital furrow. Lr thus fairly weil defined. Sz short, shallow, and on some specimens rather indis tinct. The distance between the foremost end of Sr and Sz slightly larger than the length (sag.) of Lr. On weil preserved specimens an indication of S3 can be traced in front of Sz and very close to it. No anterior pit has been observed. Occipital ring weil developed, resembling the foremost rings of the thoracic rachis, and only slightly longer (sag.) than these. Its width (tr.) about equal to the width of the glabella at S3. Occipital furrow moder ately broad, deep, widening at the junction with the dorsal furrow. Dorsal furrows of the cephalon weil defined, moderately deep. Fixigenae rather narrow. Eyes comparatively small, situated near the glabella, and rather far forwards. The distance between the posterior end of the eye and the posterior margin of the cranidium slightly larger than the shortest distance between Sr. Visual surface rather high, strongly convex. In spite of the relatively good state of preservation of the eye in one specimen no trace of lenses could be observed. Palpebral lo be narrow; no palbebral furrow. Librigenae rather feebly convex, sloping almost vertically in anterior view. In dorsal view the distal parts of the librigenae are concealed below their proximal parts. Externa! margin of the Iibrigenae broadly and evenly rounded. The genae are surrounded by a moderately wide, slightly convex border which is widest laterally. The border furrow is well-defined, fairly narrow and deep on the posterior part of the fixigena, shallower on the librigena. On the anterior part of the fixigenae the border furrow becomes almost obsolete and is visible as a faint depression which reaches the preglabellar furrow. In front of the glabella the border is rather flattened, and as it is separated from the frontal Iobe by the preglabellar furrow only, no pregla beilar field can be distinguished in this species. Facial suture evidently opisthoparian, anterior braoch more or less evenly curved, nearly parallel to the dorsal and preglabellar furrows up to the rostraJ shield. The connecting suture short, straight, and almost sagitally directed. RostraJ suture Iikewise straight, and comparatively short. RostraJ shield rather strongly convex, quadrangular, about 21/2 times broader than Iong, with a convex distal margin. Posterior braoch of the facial suture more or less straight, running in postero-lateral direction. The postero-medial corner of the Iibrigena protrudes slightly, and forms a short posteriorly directed hump at the posterior margin of the cephalon. Due to surface corrosion this hump is not clearly visible on the specimens photographed hut its general shape on som e other specimens is similar to that figured by FR. ScHMIDT (I907, PI. I, fig. I I ). Provided that this hump corresponds to a vestigial gena! spine as the present writer is inclined to suppose, the opisthoparian character of the facial suture is obvious. 44 VALDAR JAANUSSON

Fig. 2. A-D-Celmus granulatus ANG. A-anterior view of the cephalon reconstructed after the specimen RM Ar. 17990 (cf. PI. l, fig. s); B-lateral view of the cephalon reconstructed af ter the same specimen (cf. PI. I, fig. 4); C-hypostoma after a specimen from Närke belonging to S.G.U., x 2o; D-pygidium after the specimen figured on PI. I, fig. 7 (S.G.U.), x 20. E-F-Celmus barrandei VoLB. A thoracic pleura in dorsal and ventraJ views after VOLBORTH r8s8, PI. XII, fig. 5 a and 5 c.

The ornamentation of the glabella and of the genae consists of rather coarse, dosely spaced tuberdes of more or less uniform size. The diameter of the tuberdes is, as a rule, larger than the distance between them, and often twice as large. In front of the facial suture the anterior border bears closely spaced faint concentric striae which seem to continue even on the rostral shield. All examined hypostomata are samewhat incomplete and strongly corroded. ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854 45

Posterior margin of the hypostoma evenly rounded, anterior lobe rather ftattened, posterior lobe moderately convex. At the boundary between these lo bes the re exists a rather distinct, shallow middle furrow ( cf. fig. 2 C). The hypostoma is bordered posteriorly and laterally by a narrow border with up-turned outer edge. No maculae were observed. The doublure projects postero-laterally in a pair of pointed posterior wings. Anterior wings ap parently narrow, more or less triangular. Details of the hypostomal attach ment unknown. Thorax camposed of 12 segments. Rachis about as broad (tr.) as the pleural region, except for the hindmost segments where the rachis is slightly wider than the corresponding pleura. Rachis strongly convex, considerably tapering backwards, the width of the last rachial ring being only about half that of the foremost ring. Articulating half-ring about as long (med.) as the rachial ring and separated from it by a fairly deep articulating furrow. The latter widens at the junction with the moderately deep dorsal furrow. Inner part of pleura slightly convex, lying more or less harizontal in anterior view. Fulerum broadly rounded, outer part of pleura strongly bent down. No pleural furrows. The construction of the pleurae has been exhaustively described and weil illustrated already by VoLBORTH (18s8, p. 131). They are ridged and of the "type a bourrelet" according to the terminology of BARRANDE (1852, Pl. 6). On the inner part of the pleura the anterior and posterior ftanges are distinctly defined and moderately broad, bordering the central, strongly convex part of the pleura. The outer part of the pleura lacks the ftanges and is shaped as a direct continuation of the convex central part of the inner pleura (cf. fig. z E). The ftanges end abruptly at the fulerum where the whole pleura shows a distinct constriction similar to that in certain cheirurids. The ventral side of the outer part of the pleura is covered with the doublure up to the fulerum (cf. fig. 2 F) where the more or less straight inner margin of the doublure protrudes slightly in ventraJ direction. No facet is developed. The thorax possesses the same type of ornamentation as the cephalon. Pygidium small, trapezoidal (cf. fig. 2D; Pl. I, fig. 7). Width about 21/3- 21/2 greater than the length. No trace of rachis. Near the anterior margin of the pygidium the surface is elevated to form two crescent-shaped eleva tions, one on each side of the median line. The surface of the elevations slopes more or less evenly forwards and rather steeply backwards. Postero laterally of each elevation the surface of the pygidium is conspicuously depressed. The pygidium is surrounded by a narrow rounded border or edge lying only slightly higher than the inner surface of the pygidium. Distally from the border the surface of the pygidium is steeply sloping, and pro trudes into a short triangular wing at the antero-lateral corner of the py gidium. U pon the surface of the pygidium no ornamentation has been ob served. VALDAR JAANUSSON

DIMENSIONS of the specimen RM no. Ar. 17990, figured on PI. I, figs. 4-5. Measurements in mm.

r. Length of the cephalon (distance between the posterior margin of the occipital ring and the rostral s uture) ...... 6. r 2. Width of the cephalon (greatest distance between the lateral ends of the librigenae) ...... 7.8 3· Length of the glabella . . . 4·7 4· Width of the occipital ring 3·7 5· Length of the eye . . 1. 5 6. Height of the eye 1.3 7· Distance between the posterior end of the eye and the posterior margin of the cephalon ...... 2.3 8. Width of the rachis of rst thoracic segment 3·7 9· Width of the rachis of 6th thoracic segment 3.2 ro. Width of the pleural region of 6th thoracic segment . 2.9 r r. Width of the rachis of 12th thoracic segment . . . . 1.9 12. Width of the pleural region of 12th thoracic segment 1.4

DISCUSSSION.-Celmus granulatus ANG. dosely resembles Crotalurus bar randei VoLB. as described and illustrated by VoLBORTH (1858). Small diffe rences in the shape of the pygidium and in some other minor details may be due to slight inaccurateness of the drawings in VOLBORTH. To judge from VoLBORTH's figures the ornamentation of Cr. barrandei appears to be practi cally identical with that of the specimens described. The specimens of Cr. barrandei from the VOLBORTH's collection figured by FR. ScHMIDT (1907, PI. I, figs. 9-r o) differ, however, in several im portant respects from Celm:ts granulatus as well as from the VoLBORTH's figures (r 858, PI. VII, fig. r-3). The eyes are inter alia much lower, the librigenae less steeply indined, and the ornamentation consists of much finer and more scattered tuberdes. Thes� specimens are, therefore, certainly not conspecific with Cclmus granutatus. It is possible that VoLBORTH's collection indudes two different species of Celmus one described by him (r 858) and probably identical with Cdmus granulatus, and the other figured by FR. ScHMIDT (1907). But the drawings in VoLBORTH (r858) may also be incorrect in the above respects, and may actually represent the species figured by FR. SCHMIDT (1907). In this case Celmus barrandei VoLB. would be dearly different from C. granulatus. This question cannot be solved without a study of VOLBORTH's collection. There is, however, no doubt whatsoever that, as pointed out already by FR. ScHMIDT (1907), CelmY-s and Crotalurus are subjective synonyms. The specimens from the Gigas limestone of Öland determined by BoHLIN (1955) as Celmus aff. granulatus are not conspecific with C. granulatus. They p ossess inte r alia mu ch smaller tuberdes ( cf. PI. I, fig. 9). The size of the tuberdes on these specimens is on the whole equal to that in Celmus bar randei as figured by FR. ScHMIDT (r 907, PI. I, figs. 9, r o) but they are ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854 47 apparently more closely spaced. Of this species only cranidia are known so far, and more material must be studied before the species can be defined. OccuRRENCE.-The exact stratigraphical horizon is unknown for any speci men of Celmus granulatus examined hut all probably derive from the Expan sus limestone. The cotypes of ANGELIN ( 1854) were collected at Skarpåsen, Östergötland. Numerous specimens (deposited in SGU) have been collected by J. G. ANDERSSON, in all prohability from Närke, hut the exact locality is unknown.

Fam. Dimeropygidae HuP:E, 1953 Subfam. Mesotaphrasinae nov.

DrAGNOSIS.-Small opisthoparian trilobites with facial suture of levisellid type. Glabella narrowing anteriorly, two or three pairs of lateral glabellar furrows may be present. L1, if defined, relatively long. An occipital furrow present. (Hypostoma and the number of thoracic segments unknown.) Tho racic pleurae usually with a distinct pleural furrow. Pygidium small, with a well defined rachis composed of few rachial rings. DISCUSSION.-In addition to Mesotaphrasis WHITTINGTON & EviTT, 1954, and Chomatopyge WHITTINGTON & EviTT, 1954, also Clelandia CossMAN, 1902, shows the essential characters of the subfamily including the levisellid type of facial suture (cf. e.g. CLELAND 1900, HINTZE 1953). Without a first-hand knowledge of the latter genus it is, however, difficult to decide if these characters indicate a real relationship. Furthermore, attention should be paid to the general similarity between the cranidia of Mesotaphrasis and Pan archaeogonus ÖPIK, 1937. Inasmuch as the ventral cephalic sutures, thorax, and pygidium in the latter genus are unknown, its taxonornie position cannot be ascertained; neither is there, on the other hand, at present any evidence which would exclude the possibility of its belonging to Mesotaphrasinae.

Acknow ledgements The writer is indebted to the authorities of the Palaeozoological Depart ment of the State Museum of Natural History (Naturhistoriska Riksmuseet) and the Geological Survey of Sweden (Sveriges Geologiska Undersökning) for the loan of material. Dr. BIRGER BoHLIN and Dr. ToRsTEN TJERNVIK, Uppsala, have generously place d specimens collected by them at my disposal. The text figures have been prepared by Mr. E. STÅHL at the Palaeonto logical Institute, Uppsala University. VALDAR JAANUSSON

References

ANGELIN, N. P., I854: Palaeontologia Scandinavica P. I. Crustacea Formationis Tran sitionis. Lipsiae (Lundae). (2nd Edition by G. LINDSTRÖM I878.) BARRANDE, J, I852: Systeme Silurien du centre de la Boheme. I: Recherches pale ontologiques, Vol. I. Crustacis: Trilobites. Prague. -- I872: Systeme Silurien du centre de la Boheme. I : Recherches paleontologiques. Supplement au Vol. I. Trilobites, Crustaces divers et Poissons. Prague. BoHLIN, B., I955: Boring through Cambrian and Ordavieian strata at Böda Hamn, Öland. III. The Lower Ordavieian Iimestones between the Ceratopyge shale and the Platyurus limestene of Böda Hamn. Bull. Geol. Instit. of Uppsala, Vol. XXXV, No. 4; Publications from the Palaeontological Instit. of Univ. Uppsala, No. I. Uppsala. CLELAND, H. F., I900: Calciferous of the Mohawk Valley. Bull. Am. Paleont., No. I3. DALMAN, J. W., I827: Om palaeaderna eller de så kallade trilobiterna. Kungl. Vetensk. Akad. Handlingar för år I826. Stockholm. HINTZE, L. F., I953: Lower Ordovician trilobites from western Utah and eastern Nevada. Utah Geol. and Mineral. Survey, Dull. 48. Salt Lake City. HowELL, B. F., FREDERICKSON, E. A., LocHMAN, C., RAASCH, G. 0., and RASETTI, F.,

I947: Terminology for describing Cambrian trilobites . Jour. Paleontology, Vol. 2I, No. 1. Menasha, Wis. HuPE, P., I953: Classe des Trilobites. Traite de Paleontologie. Torne III. Paris. }AANUSSON, V., I954: Zur Morphologie und Taxonornie der Illaeniden. Arkiv för Mim: ralogi och Geologi, Bd. I, Nr. 20. Stockholm (printed in Uppsala). ÖPIK, A., I937: Trilobiten aus Estland. Acta et Comm. Univ. Tartuensis AXXXII; Publ. Geol. Inst. Univ. Tartu No. 52. Tartu. RASETTI, F., I 952: Ventrat cephalic sutures in Cambrian trilobites. Am. Jour. Sci., Vol. 250. I954: Phylogeny of the Cambrian trilobite fami ly Catillicephalidae and the ontogeny of Welleraspis. Jour. Paleontology, Vol. 28, No. 5· REED, F. R. C., I904: The Lower Palaeozoic trilobites of the Girvan district, Ayrshire. Part II. The Palaeontographical Society, Vol. for I904. London. RICHTER, R., I9I2: Beiträge zur Kenntnis devonischer Trilobiten· I. Die Gattung De chenella und einige verwandte Formen. Abhandl. d. Senckenb. Naturf. Gesellsch., Bd. 31. Frankfurt a. M. RICHTER, R. and RICHTER, E., I926: Die Trilobiten des Oberdevons. Beiträge zur Kenntnis devonischer Trilobiten IV. Abhandl. d. Preussischen Geol. Landesanstalt, Neue Folge, H. 99· Berlin. I 940: Die Saukianda-Stufe von Andalusien, eine frem de Fauna im europäischen Ober-Kambrium. Abhandlungen der Senckenbergischen Naiurforschenden Gesell schaft, Abh. 450. Frankfurt a. M. Ross, R. J. }R., I953: Additional Garden City (early Ordovician) trilobites. Jour. Paleontology, Vol. 27, No. 5· SALTER, J. W., I864: A monograph of the British trilobites from the Cambrian, Silu rian, and Devonian formations. Palaeontographical Society, Vol. for I862. London. ScHMIDT, FR., I 88I : Revision der ostbaltischen silurischen Trilobiten nebst geogno stischer Ubersicht des ostbaltischen Silurgebiets. Abt. I Phacopide:1, Cheiruriden und Encrinuriden. Memoires de l'Acad. Imp. des Sciences de St.-Petersbourg, VII Ser., T. XXX, No. I. St.-Petersbourg. I907: Revision der ostbaltischen silmisehen Trilobiten, Abt. VI. Ibid., VII Ser., Vol. XX, No. 8. St.-Petersbourg. ON THE TRILOBITE GENUS CELMUS ANGELIN, 1854 49

STUBBLEFIELD, C. J., I936: Cephalic s utures and their hearing on current classifications of trilobites. Biological Review, Vol. I I. Cambridge. TJERNVIK, T., I952: Om de lägsta ordoviciska lagren i Närke. Geol. Fören. Förhand!., Bd. 74, H. I. Stockholm. ULRICH, E. 0., I 930: Ordovician trilo bites of the family Telephidae and concerned stratigraphic correlations. Proceedings U.S. Nat. Mus., Vol. 76, Art. 21. Wash ington, D.C. ULRICH, E. O. and REsSER C. E., I930: The Cambrian of the upper Mississippi valley. Part I, Trilobita; Dikelocephalinae and Osceolinae. Bulletin of the Milwaukee Public Mus., Vol. I2. VoLBORTH, A. v., I858: Uber die Crotaluren und Remopleuriden, ein Beitrag zur Kenntnis der Russischen Trilobiten. Verhandl. Russisch-Kais. Mineral. Gesellsch. zu St. Petersburg, J ah r g. I 857-I 858, St. Petersburg. WALCOTT, C. D., 1908: Cambrian geology and paleontology. No. 2.- Cambrian trilo bites. Smithsonian Miscellaneous Collections, Vol. LIII, No. I8os. Washington. WARBURG, E., I925: The trilobites of the Leptaena limestone in Dalarne. Bull. Geol. Instit. Upsala, Vol. XVII. Upsala. -- I 939: The Swedish Ordovician and Lower Silurian Lichidae. Kungl. Svenska Vetenskapsahad. Handlingar, Ser. 3, Bd. I7, N:o 4· Stockholm.

WEBER, V., 1932: Trilobites of the Turkestan. - H.H.T.II. - C.C.C.P. BcecoroaHoe reonoro-PaaBen;o•moe 06"hen�meuHe. Moskva-Leningrad. WHITTINGTON, H. B. and EVITT, W. R., r 954: Silicified Middle Ordovician trilobites. Geol. Soc. America, Memoir 59· Baltimore, Md.

Plate I

All specimens whitened with ammonium chloride. Photographed by Mr. N. HJORTH, slightly retouched by Mr. E. STÅHL.

Gelmus granulatus ANGELIN, r854

r-J. LECTOTYPE.-One of the cotypes of ANGELIN (1854-). RM no. Ar. 17989. X 6. Öste' götbnd, Skarpåsen. Probably Expansus limestone.

4-5. The se:ond of the cotypes of ANGELIN (1854). RM no. Ar. 17990. X 6. Same locality and horizon. 6. Cranidium and 4 foremost thoracic segments in dorsal view. SGU. X 6. Närke. Probably Expansus limestone. Coli. J. G. ANDERSSON. 7. Pygidium in dorsal view. SGU. X ro. The same collection as the fig. 6.

Gelmus aff. granulatus ANGELIN, r854

8. Interna! mold of a glabella in dorsal view. RM no. Ar. 22985. X 6. Öland, Gässlunda. Red "Raniccps" limestone. 9· Glabella in dorsal view. UM no. Öl. r 3 r. X 6. Öland, exposure in a ditch immediately N of Långalvaret, parish of Böda (locality L in BOHLIN 1955, p. rz8). Gigas limestone. Coli. JAANussoN & MARTNA 1952. Bull. Geol. Inst., Upsala. Vol. XXXVI: z PI. I